- Email: [email protected]
G polymorphisms and endometriosis risk: a meta-analysis of case–control studies
European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Association of GSTP1 313A/G polymorphisms and endometriosis risk: a meta-analysis of case–control studies Xu Chen 1, Yulan Yan 1, Ping Li, Zheng Yang, Lingyan Qin, Wuning Mo * Department of Clinical Laboratory, First Affiliated Hospital of Guangxi Medical University, Nanning 530021, Guangxi, People’s Republic of China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 4 May 2013 Received in revised form 12 August 2013 Accepted 2 October 2013
Objectives: In view of the controversies surrounding the association of glutathione S-transferases (GST) P1 with endometriosis, a meta-analysis of GSTP1 313A/G polymorphism with endometriosis risk was performed. Study design: The relevant studies were identified through a search of PubMed, Excerpta Medica Database (Embase), Elsevier Science Direct and Chinese Biomedical Literature Database (CBM) until March 2013. The association between GSTP1 313A/G polymorphism and endometriosis risk was pooled by odds ratios (ORs) together with their 95% confidence intervals (95% CIs). Results: A total of eight case–control studies were eventually identified. We found that GSTP1-313A/G polymorphism was not associated with endometriosis risk in the overall population (A vs. G: OR = 1.02, 95% CI = 0.97–1.07, P = 0.511; AA vs. GG: OR = 1.02, 95% CI = 0.98–1.06, P = 0.359; GA vs. GG: OR = 1.03, 95% CI = 0.98–1.08, P = 0.299; AA vs. GA/GG: OR = 1.01, 95% CI = 0.96–1.07, P = 0.621; AA/GA vs. GG: OR = 1.00, 95% CI = 0.97–1.03, P = 0.972). In the sub-group analysis based on ethnicity, a significant association was found in Caucasians under the recessive model (AA vs. GA/GG: OR = 1.28, 95% CI = 1.08– 1.53, P = 0.006). Conclusions: GSTP1 313A/G polymorphism may not be associated with endometriosis risk, while the observed increase in risk of endometriosis may be due to small-study bias. Considering the limited sample size and ethnicity included in our meta-analysis, an updated meta-analysis will be urgently needed when further larger and well-designed studies are published. Crown Copyright ß 2013 Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: GSTP1 Endometriosis Polymorphism Meta-analysis
1. Introduction Endometriosis is a common inflammatory disease that is characterized by the presence of endometrial tissue outside the uterine cavity. In industrialized countries, the incidence of endometriosis has been reported to be 6–10% of the general female population, and the frequency is 35–50% in women with pain or infertility [1]. Women’s quality of life is influenced negatively by the pain, infertility and economic burden of the disease but the exact etiology of endometriosis remains unclear. Several hypotheses have been raised to explain the origin of endometriosis. Of these, retrograde menstruation is the most
* Corresponding author at: Department of Clinical Laboratory, First Affiliated Hospital of Guangxi Medical University, 22 Shuangyong Road, Nanning 530021, People’s Republic of China. Tel.: +86 13878801998. E-mail addresses: [email protected] (X. Chen), [email protected] (Y. Yan), [email protected] (P. Li), [email protected] (Z. Yang), [email protected] (L. Qin), [email protected] (W. Mo). 1 These authors contributed equally to this work so that they should be considered as the co-first author.
widely accepted theory, but still fails to explain the etiopathogenesis of this disease considering the high prevalence of this event (90%), suggesting the involvement of additional genetic or nongenetic risk factors [2]. A great deal of evidence has indicated that endometriosis is likely to be a multifactorial disease. Several factors such as immunologic, hormonal, environmental, and genetic factors may confer susceptibility to endometriosis [1,3]. It has been reported that the first-degree relatives of women with endometriosis tend to have a 5–8-fold increase in risk of occurrence of this disease [4]. About a 7-fold higher incidence was seen in relatives of women with endometriosis compared with women without a family history [5,6]. All these studies support the view that genetic factors and heritable tendencies play an important role in the development of endometriosis. Because dioxins (dioxin and dioxin-like chemicals) have an adverse impact on growth factors, cytokines, hormones and the immune system, exposure to dioxins has been proposed to lead to the development of endometriosis. Thus, polymorphisms of the detoxification enzymes which cause a lack of function or decreased function of these enzymes may contribute to endometriosis. Like other members of the glutathione-s-transferase (GST) family,
0301-2115/$ – see front matter . Crown Copyright ß 2013 Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejogrb.2013.10.005
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
GSTP1, which locates at 11q13, participates in the detoxification of electrophilic compounds by glutathione conjugation [7]. Both exon 5 and exon 6 are related to the polymorphism and result in amino acid substitutions, but only exon 5 is linked to enzymatic activity. The polymorphism of exon 5 happens at the 313 site, where an adenosine-to-guanidine (A–G) transition leads to an Ile-to-Val substitution. It has been confirmed that different GSTs have different conversion efficiency and enantio-selectivity for various toxic products [8,9]. A number of studies have reported on the association of GSTM1 and GSTT1 with endometriosis, but studies on the association between GSTP1 and the endometriosis are rare. Furthermore, it has been reported that GSTP1 is a key substance which plays an important role in biotransformation of certain environmental pollutants [10]. This suggests that GSTP1 may be more important than other GST in the pathogenesis of endometriosis. For this reason, several molecular epidemiological studies have been conducted to evaluate the association of GSTP1 313A/G polymorphism with endometriosis risk, but the previous results remain controversial rather than consistent [11–18]. Therefore, the aim of our study is to derive a more precise estimation of the association between GSTP1 313A/G polymorphism and endometriosis risk by performing a meta-analysis of all eligible case– control studies. The present research may have clinical impact if we could demonstrate that the GSTP1 313A/G polymorphism is significantly linked to endometriosis risk. Such a breakthrough may provide the introduction of predictive tests or preventive measures and even support the use of anti-oxidants in the medical treatment of endometriosis. 2. Materials and methods 2.1. Search strategy We conducted an extensive search for studies that examined the association between the GSTP1 313A/G polymorphisms and endometriosis risk. All eligible studies were identified by searching the PubMed, Embase, Elsevier Science Direct and Chinese Biomedical Literature Database (CBM) on March 20, 2013 using the terms as follows: (‘‘GSTP1’’ or ‘‘glutathione-stransferase P1’’) in combination with (‘‘polymorphism’’ or ‘‘polymorphisms’’ or ‘‘variant’’ or ‘‘mutation’’) in combination with (‘‘endometriosis’’) for all publications on the association between GSTP1 313A/G polymorphisms and endometriosis risk. No language or country restrictions were applied. Additional studies were identified by a hand search of the references of original studies. Review articles were also examined to find additional eligible studies. 2.2. Inclusion and exclusion criteria The following inclusion criteria were applied for selecting papers: (a) the paper must be a case–control study; (b) articles evaluating the association between the GSTP1 polymorphisms and endometriosis risk; (c) the paper had to provide sufficient data to estimate an odds ratio (OR) and 95% confidence interval (95% CI). The following exclusion criteria were applied for excluding papers: (a) review articles and studies that contained overlapping data; (b) case reports; (c) if more than one study from the same group occurred, we only recruited the later study. 2.3. Data extraction From each eligible study, the following information was extracted by two investigators independently with the standard
363
protocol: the name of first author, year of publication, country of origin, ethnicity of the population, the sample size of cases and controls, source of the control group, genotyping methods and allele and genotype distribution in cases and controls. Different ethnicities of the population were categorized as Caucasians, Asian, African or Mixed. The results were reviewed by a third investigator and disagreement was resolved by discussion. 2.4. Statistical analysis The possible association between GSTP1 313A/G polymorphisms and endometriosis risk was evaluated by OR and 95% CI under allele contrast (A vs. G), homozygote (AA vs. GG), heterozygote (GA vs. GG), recessive (AA vs. GA/GG), and dominant (AA/GA vs. GG) models. The heterogeneity between the studies was assessed by Q-statistic. If P < 0.10, between-study heterogeneity was considered to exist, and thus the random-effects model was used (DerSimonian and Laird method) [19]. Otherwise, the fixed-effects model (the Mantel-Haenszel method) was used to pool the data [20]. In addition, we also quantified the effect of heterogeneity using I2 [21]. The overall estimate of risk was calculated by the random-effects model when obvious heterogeneity existed (I2 value >50% or P < 0.10). Otherwise, the fixedeffects model was used when obvious heterogeneity was absent (I2 value <50% or P > 0.10). To explore sources of between-study heterogeneity, we did logistic meta-regression analyses. We examined the following study characteristics: ethnicity, source of control and study sample size. Sensitivity analysis was used to assess the stability of the pooled studies by sequential omission of individual studies. The sensitivity analysis was also conducted if Hardy–Weinberg equilibrium (HWE) disequilibrium existed (P < 0.05 was considered statistically significant). HWE was tested using a professional web-based program (http://ihg2.helmholtz-muenchen.de/cgibin/ hw/hwa1.pl). The funnel plot was used to visually observe the possible publication bias [22]. Begg’s and Egger’s tests were also used to statistically detect the publication bias (P < 0.05 was considered representative of statistically significant publication bias). STATA Software (version 9.0, Stata Corp) was applied to calculate all the available data for our meta-analysis. 3. Results 3.1. Study characteristics According to the search criteria, a total of nine publications met our inclusion criteria. Of these articles, one [23] was excluded because it only investigated GSTM1 and GSTT1 polymorphism with endometriosis risk. As a result, a total of eight studies were used for our meta-analysis [11–18]. The study characteristics of the included studies were presented in Table 1. In total, the eight case–control studies that examined the association between GSTP1 313A/G polymorphism and endometriosis risk consisted of 1638 endometriosis patients and 1842 controls. There were six studies of Asians [11,12,14–16,18] and two studies of Caucasians [13,17]. The genotype distribution in the controls of all the studies included in our meta-analysis was consistent with HWE (P > 0.05). 3.2. Quantitative synthesis of data Table 2 lists the main results of our meta-analysis. We found that GSTP1 313A/G polymorphism was not associated with endometriosis risk in the overall population (A vs. G: OR = 1.02, 95% CI = 0.97–1.07, P = 0.482, Fig. 1 for A vs. G; AA vs. GG: OR = 1.02, 95% CI = 0.98–1.07, P = 0.299, Fig. 2 for AA vs. GG; GA vs.
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
364
Table 1 General characteristics of studies included in the meta-analysis. First author
Year
Country
Ethnicity
Method of genotyping
Source of control
Sample size (case/control)
HWE of controls
Wang Matsuzaka Vichi Wu Jeon Parveen Ertunc Hur
2012 2012 2012 2011 2009 2010 2005 2004
China Japan Italy Taiwan South Korea India Turkey South Korea
Asia Asia Caucasians Asia Asia Asia Caucasians Asia
PCR-HRM Direct Sequencing PCR-RFLP TaqMan TaqMan PCR-RFLP PCR-HRM PCR-RFLP
HB HB PB HB HB PB PB HB
432/493 100/143 181/162 121/171 260/164 200/300 150/150 194/259
0.256 0.126 0.482 0.348 0.661 0.173 0.680 0.110
PCR-HRM, fluorescent quantitative PCR-based high resolution melting; PCR-RFLP, PCR-restriction fragment length polymorphism; HWE, Hardy–Weinberg equilibrium; HB, hospital based; and PB, population based.
Table 2 Results of meta-analysis for GSTP1 Comparison
Population
313 A/G polymorphism and endometriosis risk. N
Sample size
Test of association
Mode
Case
Control
OR
95% CI
P
Test of heterogeneity
x2
P
I2
A vs. G
Overall Asian Caucasians
8 6 2
3276 2614 662
3664 3040 624
1.02 0.98 1.14
0.97–1.07 0.96–1.01 0.97–1.34
0.482 0.182 0.115
R F R
24.95 7.51 4.59
0.001 0.186 0.032
71.9 33.4 78.2
AA vs. GG
Overall Asian Caucasians
8 6 2
1055 868 187
1208 1047 161
1.02 1.00 1.19
0.98–1.07 0.98–1.02 0.89–1.57
0.299 0.981 0.237
R F R
21.62 1.82 7.84
0.003 0.874 0.005
67.6 0 87.2
GA vs. GG
Overall Asian Caucasians
8 6 2
660 495 165
745 555 190
1.04 1.01 1.11
0.98–1.10 0.96–1.05 0.85–1.44
0.249 0.721 0.434
R F R
13.96 0.38 8.31
0.052 0.996 0.004
49.9 0 88.0
AA vs. GA/GG
Overall Asian Caucasians
8 6 2
1638 1307 331
1842 1530 312
1.01 0.97 1.28
0.96–1.07 0.92–1.03 1.08–1.53
0.621 0.295 0.006
R F F
14.80 4.76 1.51
0.039 0.446 0.220
52.7 0 33.6
AA/GA vs. GG
Overall Asian Caucasians
8 6 2
1638 1307 331
1842 1530 312
1.01 1.00 1.08
0.99–1.04 0.99–1.02 0.93–1.24
0.276 0.878 0.317
R F R
18.53 1.25 8.18
0.010 0.940 0.004
62.2 0 87.8
OR, odds ratio; CI, confidence interval; F, fixed effects model; and R, random effects model.
GG: OR = 1.04, 95% CI = 0.98–1.10, P = 0.249; Fig. 3 for GA vs. GG; AA vs. GA/GG: OR = 1.01, 95% CI = 0.96–1.07, P = 0.621, Fig. 4 for AA vs. GA/GG; AA/GA vs. GG: OR = 1.01, 95% CI = 0.99–1.04, P = 0.276; Fig. 5 for AA/GA vs. GG). In the sub-group analysis according to ethnicity, our metaanalysis suggested that GSTP1 A/G polymorphism was not associated with endometriosis risk in the Asian population (Table 2). Interestingly, a significant association was found in Caucasians under recessive model (AA vs. GA/GG: OR = 1.28, 95% CI = 1.08–1.53, P = 0.006, Fig. 4 and Table 2).
3.3. Heterogeneity analysis and sensitivity analysis
Fig. 1. Meta-analysis under allelic model for the association between GSTP1 G polymorphism and endometriosis risk (A vs. G).
Fig. 2. Meta-analysis under homozygote model for the association between GSTP1 313A/G polymorphism and endometriosis risk (AA vs. GG).
[(Fig._1)TD$IG]
313A/
There was significant heterogeneity for all the genetic models in the overall population. To explore the source of heterogeneity, we assessed the dominant model (AA/GA vs. GG) by ethnicity (Asian or Caucasian), source of control (population-based or hospital-based) and sample size (400 subjects or >400 subjects). As a result, ethnicity (P = 0.007) but not source of control (P > 0.05) or sample size (P > 0.05) was found to contribute to substantial heterogeneity. However, some heterogeneity still
[(Fig._2)TD$IG]
[(Fig._3)TD$IG]
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
365
3.4. Publication bias Funnel plot and Begg’s and Egger’s tests were used to assess the publication bias of our included studies. The shape of the funnel plot did not suggest any evidence of obvious asymmetry (figures not shown). Begg’s and Egger’s tests were used to provide statistical evidence of symmetry. Similarly, the results revealed the absence of publication bias (allelic model: Begg’s test P = 0.174, Egger’s Test P = 0.057; homozygote model: Begg’s test P = 0.108, Egger’s Test P = 0.105; heterozygote model: Begg’s test P = 0.536, Egger’s Test P = 0.660; recessive model: Begg’s test P = 0.060, Egger’s test P = 0.174; dominant model: Begg’s test P = 0.174, Egger’s test P = 0.197). 4. Comments Fig. 3. Meta-analysis under heterozygote model for the association between GSTP1 313A/G polymorphism and endometriosis risk (GA vs. GG).
[(Fig._4)TD$IG]
Fig. 4. Meta-analysis under recessive model for the association between GSTP1 313A/G polymorphism and endometriosis risk (AA vs. GA/GG).
existed in the Caucasian population and none of these variables could explain the heterogeneity (P > 0.05). Sensitivity analysis was used to assess the stability of pooled studies by sequential omission of individual studies. In our meta-analysis, no single study influenced the overall results qualitatively as indicated by sensitivity analysis. This indicates the robustness and reliability of our results.
[(Fig._5)TD$IG]
Fig. 5. Meta-analysis under dominant model for the association between GSTP1 313A/G polymorphism and endometriosis risk (AA/GA vs. GG).
Extensive epidemiological, experimental, and molecular evidence indicates that several factors such as immunologic, hormonal and environmental factors play an important role in the development of endometriosis. Since dioxins (dioxin and dioxin-like chemicals) have a negative effect on growth factors, cytokines, hormones and the immune system, exposure to dioxins has been considered to lead to the development of endometriosis. The role of dioxins which leads to endometriosis is neither directly damaging the cells nor combining complexes with proteins and nucleic acids but acting on the target genes in the cell nucleus through integrating with arylhydrocarbon nuclear translocator (ARNT). The GST gene is one of the target genes which regulate human metabolism, immune, endocrine and are closely related with the occurrence of endometriosis. It has been confirmed that different GST have different conversion efficiency and enantio-selectivity for various toxic products which may lead to cancer [8,9]. GSTP1 is the most important member of GST family which participates in the detoxification of electrophilic compounds by glutathione conjugation [7]. Because only GSTP1 313 A/G polymorphism links to enzymatic activity, GSTP1 313 A/G polymorphism may confer susceptibility to cancers. Numerous studies have focused on GSTP1 313 A/G polymorphism and the risk of various cancers, including lung, breast, colorectal, bladder, pancreatic, thyroid, prostate cancer, and B-cell lymphoma [24]. Up until now, several meta-analyses have been performed on GSTP1 313 A/G polymorphism and cancers, such as colorectal cancer, prostate cancer, hepatocellular carcinoma, bladder cancer, gastric cancer, lung cancer and breast cancer [25–31]. Previous studies showed that GSTP1 313 A/G polymorphism contributes to the risk of prostate, bladder and gastric cancer [25–27], but the polymorphism was not considered to be linked with susceptibility to colorectal cancer, hepatocellular carcinoma or lung or breast cancer [28–31]. This study was performed to investigate the association between GSTP1 313 A/G polymorphism and endometriosis, a disease presenting similar biological behaviors to cancers [1,32]. GSTP1 is a key substance in the biotransformation and inactivation of certain environmental pollutant. Even more, GSTP1 may play an important role in the etiology of malignant disease [10]. All this evidence suggests that GSTP1 may be more important than other GST in the pathogenesis of endometriosis [17]. Since the GSTP1 313A/G polymorphism was identified; several studies have been conducted to evaluate the association of GSTP1 313 A/ G polymorphism with endometriosis risk, but the results remain inconclusive. Ertunc et al. concluded that GSTP1 313 A/G polymorphism might modulate the risk of endometriosis with significantly decreased risk for GSTP1 val/val and marginally increased risk for GSTP1 ile/ile [17]. Other studies, however, suggested that GSTP1 313 A/G polymorphism was not associated
366
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
with endometriosis risk [11–16,18]. To help explain the diversity of results from different studies, we conducted a meta-analysis to comprehensively explore the GSTP1 313 A/G polymorphism and endometriosis risk. To the best of our knowledge, this is the first meta-analysis of the assessment for the association between GSTP1 313 A/G polymorphism and endometriosis risk. In the present study, an obvious association was found only in Caucasians under the recessive model (AA vs. GA/GG: OR = 1.28, 95% CI = 1.08–1.53, P = 0.006, Fig. 4 and Table 2). No obvious associations were found in other genetic models in Caucasians, and two studies with small sample size (331 cases and 312 controls) investigated the Caucasians. We concluded that GSTP1 313 A/G polymorphism may not be associated with endometriosis risk, while the observed increase in risk of endometriosis may be due to small-study bias. Small-study bias in genetic association studies is not a new event when performing meta-analysis. For example, Li et al. published a meta-analysis to evaluate the association between ER-a gene PvuII T/C polymorphisms and endometriosis risk [33]. They concluded that the observed small nominal increase in risk of endometriosis associated with the ER-a PvuII T/C polymorphisms is due to small-study bias, which was similar with our results. Obvious heterogeneity among studies was observed in the overall population. Although we found that ethnicity contributed substantial heterogeneity to our results by using the method of meta-regression, some heterogeneity cannot be explained by several possible source of heterogeneity such as ethnicity, sample size and source of controls. In the present meta-analysis, the eligible studies used several genotyping methods including PCRRFLP, PCR-HRM, TaqMan and Direct Sequencing. All of these genotyping methods are accurate, rapid and commonly used in detecting gene polymorphism of molecular biology experiments. In previous meta-analyses, these genotyping methods were not considered to be the source of heterogeneity. Probably other limitations of included studies may partially contribute to the observed heterogeneity. Hence, we conducted analyses using the random effects model. Another important problem which may have a negative effect on the results is publication bias. In the present study, Funnel plot, and Begg’s and Egger’s tests were used to assess the publication bias of our included studies. Both the shape of funnel plot and statistical results did not show obvious publication bias. This suggests that the publication bias would not make substantial negative effect on our results and that results of our meta-analysis are relatively stable. Endocrine disrupting chemicals (EDCs) have the capacity to interfere with normal signalling systems. EDCs may mimic, block or modulate the synthesis, release, transport and metabolism of natural hormones [34]. Many studies have demonstrated that EDC exposure is a risk factor for women’s health and it is closely associated with many kinds of gynecological diseases, including endometriosis [35–37]. It is a pity that not all of the eligible studies included in our meta-analysis provide information about EDC exposure, so that we could not evaluate the impact of EDC exposure in endometriosis. Although comprehensive analysis was conducted to demonstrate the association between GSTP1 313A/G polymorphism and risk of endometriosis, there are still some limitations should be pointed out. First, only three of the eight included studies used controls that were population-based [13,16,17]. Other studies used hospital-based controls, which may not be representative of the general population [11,12,14,15,18]. Second, the number of studies and the number of subjects included in our meta-analysis were relatively small. Furthermore, there was no original study which investigated African or other ethnicities. Therefore, additional studies with larger sample size and providing detailed information should be conducted to assess the effect of GSTP1 313A/G
polymorphism on endometriosis risk in different ethnicities, especially in Africans. Third, as some studies included in our meta-analysis are based on unadjusted estimates, some risk factors such as age, family history and environment factors might cause confounding bias. In spite of the above shortcomings, our meta-analysis also had some advantages. First, a meta-analysis of the association of GSTP1 313A/G polymorphism on endometriosis risk is statistically more powerful than any single study. Second, a rigorous searching strategy which consisted of computer-assisted and manual searches made the eligible studies as inclusive as possible. Third, the quality of the case–control studies included in our metaanalysis was satisfactory and met our inclusion criteria, which leads to a more reliable result. Furthermore, our meta-analysis was well designed and conducted in the process of literature selection, data extraction and data analysis. In conclusion, GSTP1 313A/G polymorphism may not be associated with endometriosis risk, while the observed increase in risk of endometriosis may be due to small-study bias. Considering the limited sample size and ethnicities included in our metaanalysis, an updated meta-analysis will be urgently needed to confirm our results when further larger and well-designed studies are published. Acknowledgements This study has been supported by some students in acquisition of data and searching background information relevant to our study. We would like to thank them for their help which led to improvement of this article. References [1] Giudice LC, Kao LC. Endometriosis. Lancet 2004;364:1789–99. [2] Halme J, Hammond MG, Hulka JF, Raj SG, Talbert LM. Retrograde menstruation in healthy women and in patients with endometriosis. Obstet Gynecol 1984;64:151–4. [3] Rier SE. Environmental immune disruption: a comorbidity factor for reproduction? Fertil Steril 2008;89:e103–8. [4] Bischoff F, Simpson JL. Genetic basis of endometriosis. Ann N Y Acad Sci 2004;1034:284–99. [5] Simpson JL, Elias S, Malinak LR, Buttram Jr VC. Heritable aspects of endometriosis. I. Genetic studies. Am J Obstet Gynecol 1980;137:327–31. [6] Matalliotakis IM, Arici A, Cakmak H, Goumenou AG, Koumantakis G, Mahutte NG. Familial aggregation of endometriosis in the Yale Series. Arch Gynecol Obstet 2008;278:507–11. [7] Autrup H. Genetic polymorphisms in human xenobiotica metabolizing enzymes as susceptibility factors in toxic response. Mutat Res 2000;464:65–76. [8] Hu X, Benson PJ, Srivastava SK, et al. Glutathione S-transferases of female A/J mouse liver and fore stomach and their different induction by anticarcinogenic organosulfides from garlic. Arch Biochem Biophys 1996;336:199–214. [9] Hu X, Srivastava SK, Xia H, Awasthi YC, Singh SV. An alpha class mouse glutathione S-transferase with exceptional catalytic efficiency in the conjugation of glutathione with 7b, 8a-dihydroxy-9a, 10a-oxy-7,8,9,10-tetrahydrobenzo[a] pyrene. J Biol Chem 1996;271:32684–88. [10] Harries LW, Stubbins MJ, Forman D, Howard GC, Wolf CR. Identification of genetic polymorphisms at the glutathione S-transferase Pi locus and association with susceptibility to bladder, testicular and prostate cancer. Carcinogenesis 1997;18:641–4. [11] Wang YF, Zong LL, Mao T, et al. Relationship of polymorphisms of AhR-1661G/ A with GSTP1-313A/G and susceptibility to endometriosis. Zhonghua Fu Chan Ke Za Zhi 2012;47:522–5 [Chinese]. [12] Matsuzaka Y, Kikuti YY, Goya K, et al. Lack of an association human dioxin detoxification gene polymorphisms with endometriosis in Japanese women: results of a pilot study. Environ Health Prev Med 2012;17:512–7. [13] Vichi S, Medda E, Ingelido AM, et al. Glutathione transferase polymorphisms and risk of endometriosis associated with polychlorinated biphenyls exposure in Italian women: a gene–environment interaction. Fertil Steril 2012;97:1143–51. [14] Wu CH, Guo CY, Yang JG, et al. Polymorphisms of dioxin receptor complex components and detoxification-related genes jointly confer susceptibility to advanced-stage endometriosis in the taiwanese han population. Am J Reprod Immunol 2012;67:160–8. [15] Jeon MJ, Choi YM, Hong MA, et al. No association between the GSTP1 exon 5 polymorphism and susceptibility to advanced stage endometriosis in the Korean population. Am J Reprod Immunol 2010;63:222–6.
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367 [16] Parveen F, Faridi RM, Das V, Tripathi G, Agrawal S. Genetic association of phase I and phase II detoxification genes with recurrent miscarriages among North Indian women. Mol Hum Reprod 2010;16:207–14. [17] Ertunc D, Aban M, Tok EC, Tamer L, Arslan M, Dilek S. Glutathione-S-transferase P1 gene polymorphism and susceptibility to endometriosis. Hum Reprod 2005;20:2157–61. [18] Hur SE, Lee JY, Moon HS, Chung HW. Polymorphisms of the genes encoding the GSTM1 GSTT1 and GSTP1 in Korean women: no association with endometriosis. Mol Hum Reprod 2005;11:15–9. [19] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88. [20] Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:719–48. [21] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [22] Munafo MR, Clark TG, Flint J. Assessing publication bias in genetic association studies: evidence from a recent meta-analysis. Psychiatry Res 2004;129:39–44. [23] Aban M, Ertunc D, Tok EC, Tamer L, Arslan M, Dilek S. Modulating interaction of glutathione-S-transferase polymorphisms with smoking in endometriosis. J Reprod Med 2007;52:715–21. [24] Huang SX, Wu FX, Luo M, et al. The glutathione S-transferase P1 341C>T polymorphism and cancer risk: a meta-analysis of 28 case–control studies. PLoS ONE 2013;8:e56722. [25] Yu Z, Li Z, Cai B, et al. Association between the GSTP1 Ile105Val polymorphism and prostate cancer risk: a systematic review and meta-analysis. Tumour Biol 2013;34:1855–63. [26] Wang Z, Xue L, Chong T, Li H, Chen H, Wang Z. Quantitative assessment of the association between glutathione S-transferase P1 Ile105Val polymorphism and bladder cancer risk. Tumour Biol 2013;34:1651–7.
367
[27] Ma Y, Wei X, Han G, Xue M, Li G, Li Y. Glutathione S-transferase P1 Ile105Val polymorphism contributes to increased risk of gastric cancer in East Asians. Tumour Biol 2013;34:1737–42. [28] Tan Z, Feng M, Luo Y, et al. GSTP1 Ile105Val polymorphism and colorectal cancer risk: an updated analysis. Gene 2013;527:275–8. [29] Zhao Y, Wang Q, Deng X, Shi P, Wang Z. Quantitative assessment of the association between GSTP1 gene Ile105Val polymorphism and susceptibility to hepatocellular carcinoma. Tumour Biol 2013;34:2121–6. [30] Zhong H, Feng Y, Zheng GX, et al. A meta-analysis of the association between glutathione S-transferaseP1 gene polymorphism and the risk of adenocarcinomas of lung cancer. Cancer Biomark 2013;13:29–35. [31] Liu JJ, Liu JL, Zhang X, Xie L, Zeng J. A meta-analysis of the association of glutathione S-transferase P1 gene polymorphism with the susceptibility of breast cancer. Mol Biol Rep 2013;40:3203–12. [32] Brosens I, Benagiano G, Endometriosis. a modern syndrome. Ind J Med Res 2011;133:581–93. [33] Li Y, Liu F, Tan SQ, Wang Y, Li SW. Estrogen receptor-alpha gene PvuII (T/C) and XbaI (A/G) polymorphisms and endometriosis risk: a meta-analysis. Gene 2012;508:41–8. [34] Caserta D, Maranghi L, Mantovani A, Marci R, Maranghi F, Moscarini M. Impact of endocrine disruptor chemicals in gynaecology. Hum Reprod Update 2008;14:59–72. [35] Caserta D, Bordi G, Ciardo F, et al. The influence of endocrine disruptors in a selected population of infertile women. Gynecol Endocrinol 2013;29:444–7. [36] Caserta D, Mantovani A, Marci R, et al. Environment and women’s reproductive health. Hum Reprod Update 2011;17:418–33. [37] Gentilini D, Perino A, Vigano` P, et al. Gene expression profiling of peripheral blood mononuclear cells in endometriosis identifies genes altered in nongynaecologic chronic inflammatory diseases. Hum Reprod 2011;26:3109–17.
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Association of GSTP1 313A/G polymorphisms and endometriosis risk: a meta-analysis of case–control studies Xu Chen 1, Yulan Yan 1, Ping Li, Zheng Yang, Lingyan Qin, Wuning Mo * Department of Clinical Laboratory, First Affiliated Hospital of Guangxi Medical University, Nanning 530021, Guangxi, People’s Republic of China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 4 May 2013 Received in revised form 12 August 2013 Accepted 2 October 2013
Objectives: In view of the controversies surrounding the association of glutathione S-transferases (GST) P1 with endometriosis, a meta-analysis of GSTP1 313A/G polymorphism with endometriosis risk was performed. Study design: The relevant studies were identified through a search of PubMed, Excerpta Medica Database (Embase), Elsevier Science Direct and Chinese Biomedical Literature Database (CBM) until March 2013. The association between GSTP1 313A/G polymorphism and endometriosis risk was pooled by odds ratios (ORs) together with their 95% confidence intervals (95% CIs). Results: A total of eight case–control studies were eventually identified. We found that GSTP1-313A/G polymorphism was not associated with endometriosis risk in the overall population (A vs. G: OR = 1.02, 95% CI = 0.97–1.07, P = 0.511; AA vs. GG: OR = 1.02, 95% CI = 0.98–1.06, P = 0.359; GA vs. GG: OR = 1.03, 95% CI = 0.98–1.08, P = 0.299; AA vs. GA/GG: OR = 1.01, 95% CI = 0.96–1.07, P = 0.621; AA/GA vs. GG: OR = 1.00, 95% CI = 0.97–1.03, P = 0.972). In the sub-group analysis based on ethnicity, a significant association was found in Caucasians under the recessive model (AA vs. GA/GG: OR = 1.28, 95% CI = 1.08– 1.53, P = 0.006). Conclusions: GSTP1 313A/G polymorphism may not be associated with endometriosis risk, while the observed increase in risk of endometriosis may be due to small-study bias. Considering the limited sample size and ethnicity included in our meta-analysis, an updated meta-analysis will be urgently needed when further larger and well-designed studies are published. Crown Copyright ß 2013 Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: GSTP1 Endometriosis Polymorphism Meta-analysis
1. Introduction Endometriosis is a common inflammatory disease that is characterized by the presence of endometrial tissue outside the uterine cavity. In industrialized countries, the incidence of endometriosis has been reported to be 6–10% of the general female population, and the frequency is 35–50% in women with pain or infertility [1]. Women’s quality of life is influenced negatively by the pain, infertility and economic burden of the disease but the exact etiology of endometriosis remains unclear. Several hypotheses have been raised to explain the origin of endometriosis. Of these, retrograde menstruation is the most
* Corresponding author at: Department of Clinical Laboratory, First Affiliated Hospital of Guangxi Medical University, 22 Shuangyong Road, Nanning 530021, People’s Republic of China. Tel.: +86 13878801998. E-mail addresses: [email protected] (X. Chen), [email protected] (Y. Yan), [email protected] (P. Li), [email protected] (Z. Yang), [email protected] (L. Qin), [email protected] (W. Mo). 1 These authors contributed equally to this work so that they should be considered as the co-first author.
widely accepted theory, but still fails to explain the etiopathogenesis of this disease considering the high prevalence of this event (90%), suggesting the involvement of additional genetic or nongenetic risk factors [2]. A great deal of evidence has indicated that endometriosis is likely to be a multifactorial disease. Several factors such as immunologic, hormonal, environmental, and genetic factors may confer susceptibility to endometriosis [1,3]. It has been reported that the first-degree relatives of women with endometriosis tend to have a 5–8-fold increase in risk of occurrence of this disease [4]. About a 7-fold higher incidence was seen in relatives of women with endometriosis compared with women without a family history [5,6]. All these studies support the view that genetic factors and heritable tendencies play an important role in the development of endometriosis. Because dioxins (dioxin and dioxin-like chemicals) have an adverse impact on growth factors, cytokines, hormones and the immune system, exposure to dioxins has been proposed to lead to the development of endometriosis. Thus, polymorphisms of the detoxification enzymes which cause a lack of function or decreased function of these enzymes may contribute to endometriosis. Like other members of the glutathione-s-transferase (GST) family,
0301-2115/$ – see front matter . Crown Copyright ß 2013 Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejogrb.2013.10.005
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
GSTP1, which locates at 11q13, participates in the detoxification of electrophilic compounds by glutathione conjugation [7]. Both exon 5 and exon 6 are related to the polymorphism and result in amino acid substitutions, but only exon 5 is linked to enzymatic activity. The polymorphism of exon 5 happens at the 313 site, where an adenosine-to-guanidine (A–G) transition leads to an Ile-to-Val substitution. It has been confirmed that different GSTs have different conversion efficiency and enantio-selectivity for various toxic products [8,9]. A number of studies have reported on the association of GSTM1 and GSTT1 with endometriosis, but studies on the association between GSTP1 and the endometriosis are rare. Furthermore, it has been reported that GSTP1 is a key substance which plays an important role in biotransformation of certain environmental pollutants [10]. This suggests that GSTP1 may be more important than other GST in the pathogenesis of endometriosis. For this reason, several molecular epidemiological studies have been conducted to evaluate the association of GSTP1 313A/G polymorphism with endometriosis risk, but the previous results remain controversial rather than consistent [11–18]. Therefore, the aim of our study is to derive a more precise estimation of the association between GSTP1 313A/G polymorphism and endometriosis risk by performing a meta-analysis of all eligible case– control studies. The present research may have clinical impact if we could demonstrate that the GSTP1 313A/G polymorphism is significantly linked to endometriosis risk. Such a breakthrough may provide the introduction of predictive tests or preventive measures and even support the use of anti-oxidants in the medical treatment of endometriosis. 2. Materials and methods 2.1. Search strategy We conducted an extensive search for studies that examined the association between the GSTP1 313A/G polymorphisms and endometriosis risk. All eligible studies were identified by searching the PubMed, Embase, Elsevier Science Direct and Chinese Biomedical Literature Database (CBM) on March 20, 2013 using the terms as follows: (‘‘GSTP1’’ or ‘‘glutathione-stransferase P1’’) in combination with (‘‘polymorphism’’ or ‘‘polymorphisms’’ or ‘‘variant’’ or ‘‘mutation’’) in combination with (‘‘endometriosis’’) for all publications on the association between GSTP1 313A/G polymorphisms and endometriosis risk. No language or country restrictions were applied. Additional studies were identified by a hand search of the references of original studies. Review articles were also examined to find additional eligible studies. 2.2. Inclusion and exclusion criteria The following inclusion criteria were applied for selecting papers: (a) the paper must be a case–control study; (b) articles evaluating the association between the GSTP1 polymorphisms and endometriosis risk; (c) the paper had to provide sufficient data to estimate an odds ratio (OR) and 95% confidence interval (95% CI). The following exclusion criteria were applied for excluding papers: (a) review articles and studies that contained overlapping data; (b) case reports; (c) if more than one study from the same group occurred, we only recruited the later study. 2.3. Data extraction From each eligible study, the following information was extracted by two investigators independently with the standard
363
protocol: the name of first author, year of publication, country of origin, ethnicity of the population, the sample size of cases and controls, source of the control group, genotyping methods and allele and genotype distribution in cases and controls. Different ethnicities of the population were categorized as Caucasians, Asian, African or Mixed. The results were reviewed by a third investigator and disagreement was resolved by discussion. 2.4. Statistical analysis The possible association between GSTP1 313A/G polymorphisms and endometriosis risk was evaluated by OR and 95% CI under allele contrast (A vs. G), homozygote (AA vs. GG), heterozygote (GA vs. GG), recessive (AA vs. GA/GG), and dominant (AA/GA vs. GG) models. The heterogeneity between the studies was assessed by Q-statistic. If P < 0.10, between-study heterogeneity was considered to exist, and thus the random-effects model was used (DerSimonian and Laird method) [19]. Otherwise, the fixed-effects model (the Mantel-Haenszel method) was used to pool the data [20]. In addition, we also quantified the effect of heterogeneity using I2 [21]. The overall estimate of risk was calculated by the random-effects model when obvious heterogeneity existed (I2 value >50% or P < 0.10). Otherwise, the fixedeffects model was used when obvious heterogeneity was absent (I2 value <50% or P > 0.10). To explore sources of between-study heterogeneity, we did logistic meta-regression analyses. We examined the following study characteristics: ethnicity, source of control and study sample size. Sensitivity analysis was used to assess the stability of the pooled studies by sequential omission of individual studies. The sensitivity analysis was also conducted if Hardy–Weinberg equilibrium (HWE) disequilibrium existed (P < 0.05 was considered statistically significant). HWE was tested using a professional web-based program (http://ihg2.helmholtz-muenchen.de/cgibin/ hw/hwa1.pl). The funnel plot was used to visually observe the possible publication bias [22]. Begg’s and Egger’s tests were also used to statistically detect the publication bias (P < 0.05 was considered representative of statistically significant publication bias). STATA Software (version 9.0, Stata Corp) was applied to calculate all the available data for our meta-analysis. 3. Results 3.1. Study characteristics According to the search criteria, a total of nine publications met our inclusion criteria. Of these articles, one [23] was excluded because it only investigated GSTM1 and GSTT1 polymorphism with endometriosis risk. As a result, a total of eight studies were used for our meta-analysis [11–18]. The study characteristics of the included studies were presented in Table 1. In total, the eight case–control studies that examined the association between GSTP1 313A/G polymorphism and endometriosis risk consisted of 1638 endometriosis patients and 1842 controls. There were six studies of Asians [11,12,14–16,18] and two studies of Caucasians [13,17]. The genotype distribution in the controls of all the studies included in our meta-analysis was consistent with HWE (P > 0.05). 3.2. Quantitative synthesis of data Table 2 lists the main results of our meta-analysis. We found that GSTP1 313A/G polymorphism was not associated with endometriosis risk in the overall population (A vs. G: OR = 1.02, 95% CI = 0.97–1.07, P = 0.482, Fig. 1 for A vs. G; AA vs. GG: OR = 1.02, 95% CI = 0.98–1.07, P = 0.299, Fig. 2 for AA vs. GG; GA vs.
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
364
Table 1 General characteristics of studies included in the meta-analysis. First author
Year
Country
Ethnicity
Method of genotyping
Source of control
Sample size (case/control)
HWE of controls
Wang Matsuzaka Vichi Wu Jeon Parveen Ertunc Hur
2012 2012 2012 2011 2009 2010 2005 2004
China Japan Italy Taiwan South Korea India Turkey South Korea
Asia Asia Caucasians Asia Asia Asia Caucasians Asia
PCR-HRM Direct Sequencing PCR-RFLP TaqMan TaqMan PCR-RFLP PCR-HRM PCR-RFLP
HB HB PB HB HB PB PB HB
432/493 100/143 181/162 121/171 260/164 200/300 150/150 194/259
0.256 0.126 0.482 0.348 0.661 0.173 0.680 0.110
PCR-HRM, fluorescent quantitative PCR-based high resolution melting; PCR-RFLP, PCR-restriction fragment length polymorphism; HWE, Hardy–Weinberg equilibrium; HB, hospital based; and PB, population based.
Table 2 Results of meta-analysis for GSTP1 Comparison
Population
313 A/G polymorphism and endometriosis risk. N
Sample size
Test of association
Mode
Case
Control
OR
95% CI
P
Test of heterogeneity
x2
P
I2
A vs. G
Overall Asian Caucasians
8 6 2
3276 2614 662
3664 3040 624
1.02 0.98 1.14
0.97–1.07 0.96–1.01 0.97–1.34
0.482 0.182 0.115
R F R
24.95 7.51 4.59
0.001 0.186 0.032
71.9 33.4 78.2
AA vs. GG
Overall Asian Caucasians
8 6 2
1055 868 187
1208 1047 161
1.02 1.00 1.19
0.98–1.07 0.98–1.02 0.89–1.57
0.299 0.981 0.237
R F R
21.62 1.82 7.84
0.003 0.874 0.005
67.6 0 87.2
GA vs. GG
Overall Asian Caucasians
8 6 2
660 495 165
745 555 190
1.04 1.01 1.11
0.98–1.10 0.96–1.05 0.85–1.44
0.249 0.721 0.434
R F R
13.96 0.38 8.31
0.052 0.996 0.004
49.9 0 88.0
AA vs. GA/GG
Overall Asian Caucasians
8 6 2
1638 1307 331
1842 1530 312
1.01 0.97 1.28
0.96–1.07 0.92–1.03 1.08–1.53
0.621 0.295 0.006
R F F
14.80 4.76 1.51
0.039 0.446 0.220
52.7 0 33.6
AA/GA vs. GG
Overall Asian Caucasians
8 6 2
1638 1307 331
1842 1530 312
1.01 1.00 1.08
0.99–1.04 0.99–1.02 0.93–1.24
0.276 0.878 0.317
R F R
18.53 1.25 8.18
0.010 0.940 0.004
62.2 0 87.8
OR, odds ratio; CI, confidence interval; F, fixed effects model; and R, random effects model.
GG: OR = 1.04, 95% CI = 0.98–1.10, P = 0.249; Fig. 3 for GA vs. GG; AA vs. GA/GG: OR = 1.01, 95% CI = 0.96–1.07, P = 0.621, Fig. 4 for AA vs. GA/GG; AA/GA vs. GG: OR = 1.01, 95% CI = 0.99–1.04, P = 0.276; Fig. 5 for AA/GA vs. GG). In the sub-group analysis according to ethnicity, our metaanalysis suggested that GSTP1 A/G polymorphism was not associated with endometriosis risk in the Asian population (Table 2). Interestingly, a significant association was found in Caucasians under recessive model (AA vs. GA/GG: OR = 1.28, 95% CI = 1.08–1.53, P = 0.006, Fig. 4 and Table 2).
3.3. Heterogeneity analysis and sensitivity analysis
Fig. 1. Meta-analysis under allelic model for the association between GSTP1 G polymorphism and endometriosis risk (A vs. G).
Fig. 2. Meta-analysis under homozygote model for the association between GSTP1 313A/G polymorphism and endometriosis risk (AA vs. GG).
[(Fig._1)TD$IG]
313A/
There was significant heterogeneity for all the genetic models in the overall population. To explore the source of heterogeneity, we assessed the dominant model (AA/GA vs. GG) by ethnicity (Asian or Caucasian), source of control (population-based or hospital-based) and sample size (400 subjects or >400 subjects). As a result, ethnicity (P = 0.007) but not source of control (P > 0.05) or sample size (P > 0.05) was found to contribute to substantial heterogeneity. However, some heterogeneity still
[(Fig._2)TD$IG]
[(Fig._3)TD$IG]
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
365
3.4. Publication bias Funnel plot and Begg’s and Egger’s tests were used to assess the publication bias of our included studies. The shape of the funnel plot did not suggest any evidence of obvious asymmetry (figures not shown). Begg’s and Egger’s tests were used to provide statistical evidence of symmetry. Similarly, the results revealed the absence of publication bias (allelic model: Begg’s test P = 0.174, Egger’s Test P = 0.057; homozygote model: Begg’s test P = 0.108, Egger’s Test P = 0.105; heterozygote model: Begg’s test P = 0.536, Egger’s Test P = 0.660; recessive model: Begg’s test P = 0.060, Egger’s test P = 0.174; dominant model: Begg’s test P = 0.174, Egger’s test P = 0.197). 4. Comments Fig. 3. Meta-analysis under heterozygote model for the association between GSTP1 313A/G polymorphism and endometriosis risk (GA vs. GG).
[(Fig._4)TD$IG]
Fig. 4. Meta-analysis under recessive model for the association between GSTP1 313A/G polymorphism and endometriosis risk (AA vs. GA/GG).
existed in the Caucasian population and none of these variables could explain the heterogeneity (P > 0.05). Sensitivity analysis was used to assess the stability of pooled studies by sequential omission of individual studies. In our meta-analysis, no single study influenced the overall results qualitatively as indicated by sensitivity analysis. This indicates the robustness and reliability of our results.
[(Fig._5)TD$IG]
Fig. 5. Meta-analysis under dominant model for the association between GSTP1 313A/G polymorphism and endometriosis risk (AA/GA vs. GG).
Extensive epidemiological, experimental, and molecular evidence indicates that several factors such as immunologic, hormonal and environmental factors play an important role in the development of endometriosis. Since dioxins (dioxin and dioxin-like chemicals) have a negative effect on growth factors, cytokines, hormones and the immune system, exposure to dioxins has been considered to lead to the development of endometriosis. The role of dioxins which leads to endometriosis is neither directly damaging the cells nor combining complexes with proteins and nucleic acids but acting on the target genes in the cell nucleus through integrating with arylhydrocarbon nuclear translocator (ARNT). The GST gene is one of the target genes which regulate human metabolism, immune, endocrine and are closely related with the occurrence of endometriosis. It has been confirmed that different GST have different conversion efficiency and enantio-selectivity for various toxic products which may lead to cancer [8,9]. GSTP1 is the most important member of GST family which participates in the detoxification of electrophilic compounds by glutathione conjugation [7]. Because only GSTP1 313 A/G polymorphism links to enzymatic activity, GSTP1 313 A/G polymorphism may confer susceptibility to cancers. Numerous studies have focused on GSTP1 313 A/G polymorphism and the risk of various cancers, including lung, breast, colorectal, bladder, pancreatic, thyroid, prostate cancer, and B-cell lymphoma [24]. Up until now, several meta-analyses have been performed on GSTP1 313 A/G polymorphism and cancers, such as colorectal cancer, prostate cancer, hepatocellular carcinoma, bladder cancer, gastric cancer, lung cancer and breast cancer [25–31]. Previous studies showed that GSTP1 313 A/G polymorphism contributes to the risk of prostate, bladder and gastric cancer [25–27], but the polymorphism was not considered to be linked with susceptibility to colorectal cancer, hepatocellular carcinoma or lung or breast cancer [28–31]. This study was performed to investigate the association between GSTP1 313 A/G polymorphism and endometriosis, a disease presenting similar biological behaviors to cancers [1,32]. GSTP1 is a key substance in the biotransformation and inactivation of certain environmental pollutant. Even more, GSTP1 may play an important role in the etiology of malignant disease [10]. All this evidence suggests that GSTP1 may be more important than other GST in the pathogenesis of endometriosis [17]. Since the GSTP1 313A/G polymorphism was identified; several studies have been conducted to evaluate the association of GSTP1 313 A/ G polymorphism with endometriosis risk, but the results remain inconclusive. Ertunc et al. concluded that GSTP1 313 A/G polymorphism might modulate the risk of endometriosis with significantly decreased risk for GSTP1 val/val and marginally increased risk for GSTP1 ile/ile [17]. Other studies, however, suggested that GSTP1 313 A/G polymorphism was not associated
366
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367
with endometriosis risk [11–16,18]. To help explain the diversity of results from different studies, we conducted a meta-analysis to comprehensively explore the GSTP1 313 A/G polymorphism and endometriosis risk. To the best of our knowledge, this is the first meta-analysis of the assessment for the association between GSTP1 313 A/G polymorphism and endometriosis risk. In the present study, an obvious association was found only in Caucasians under the recessive model (AA vs. GA/GG: OR = 1.28, 95% CI = 1.08–1.53, P = 0.006, Fig. 4 and Table 2). No obvious associations were found in other genetic models in Caucasians, and two studies with small sample size (331 cases and 312 controls) investigated the Caucasians. We concluded that GSTP1 313 A/G polymorphism may not be associated with endometriosis risk, while the observed increase in risk of endometriosis may be due to small-study bias. Small-study bias in genetic association studies is not a new event when performing meta-analysis. For example, Li et al. published a meta-analysis to evaluate the association between ER-a gene PvuII T/C polymorphisms and endometriosis risk [33]. They concluded that the observed small nominal increase in risk of endometriosis associated with the ER-a PvuII T/C polymorphisms is due to small-study bias, which was similar with our results. Obvious heterogeneity among studies was observed in the overall population. Although we found that ethnicity contributed substantial heterogeneity to our results by using the method of meta-regression, some heterogeneity cannot be explained by several possible source of heterogeneity such as ethnicity, sample size and source of controls. In the present meta-analysis, the eligible studies used several genotyping methods including PCRRFLP, PCR-HRM, TaqMan and Direct Sequencing. All of these genotyping methods are accurate, rapid and commonly used in detecting gene polymorphism of molecular biology experiments. In previous meta-analyses, these genotyping methods were not considered to be the source of heterogeneity. Probably other limitations of included studies may partially contribute to the observed heterogeneity. Hence, we conducted analyses using the random effects model. Another important problem which may have a negative effect on the results is publication bias. In the present study, Funnel plot, and Begg’s and Egger’s tests were used to assess the publication bias of our included studies. Both the shape of funnel plot and statistical results did not show obvious publication bias. This suggests that the publication bias would not make substantial negative effect on our results and that results of our meta-analysis are relatively stable. Endocrine disrupting chemicals (EDCs) have the capacity to interfere with normal signalling systems. EDCs may mimic, block or modulate the synthesis, release, transport and metabolism of natural hormones [34]. Many studies have demonstrated that EDC exposure is a risk factor for women’s health and it is closely associated with many kinds of gynecological diseases, including endometriosis [35–37]. It is a pity that not all of the eligible studies included in our meta-analysis provide information about EDC exposure, so that we could not evaluate the impact of EDC exposure in endometriosis. Although comprehensive analysis was conducted to demonstrate the association between GSTP1 313A/G polymorphism and risk of endometriosis, there are still some limitations should be pointed out. First, only three of the eight included studies used controls that were population-based [13,16,17]. Other studies used hospital-based controls, which may not be representative of the general population [11,12,14,15,18]. Second, the number of studies and the number of subjects included in our meta-analysis were relatively small. Furthermore, there was no original study which investigated African or other ethnicities. Therefore, additional studies with larger sample size and providing detailed information should be conducted to assess the effect of GSTP1 313A/G
polymorphism on endometriosis risk in different ethnicities, especially in Africans. Third, as some studies included in our meta-analysis are based on unadjusted estimates, some risk factors such as age, family history and environment factors might cause confounding bias. In spite of the above shortcomings, our meta-analysis also had some advantages. First, a meta-analysis of the association of GSTP1 313A/G polymorphism on endometriosis risk is statistically more powerful than any single study. Second, a rigorous searching strategy which consisted of computer-assisted and manual searches made the eligible studies as inclusive as possible. Third, the quality of the case–control studies included in our metaanalysis was satisfactory and met our inclusion criteria, which leads to a more reliable result. Furthermore, our meta-analysis was well designed and conducted in the process of literature selection, data extraction and data analysis. In conclusion, GSTP1 313A/G polymorphism may not be associated with endometriosis risk, while the observed increase in risk of endometriosis may be due to small-study bias. Considering the limited sample size and ethnicities included in our metaanalysis, an updated meta-analysis will be urgently needed to confirm our results when further larger and well-designed studies are published. Acknowledgements This study has been supported by some students in acquisition of data and searching background information relevant to our study. We would like to thank them for their help which led to improvement of this article. References [1] Giudice LC, Kao LC. Endometriosis. Lancet 2004;364:1789–99. [2] Halme J, Hammond MG, Hulka JF, Raj SG, Talbert LM. Retrograde menstruation in healthy women and in patients with endometriosis. Obstet Gynecol 1984;64:151–4. [3] Rier SE. Environmental immune disruption: a comorbidity factor for reproduction? Fertil Steril 2008;89:e103–8. [4] Bischoff F, Simpson JL. Genetic basis of endometriosis. Ann N Y Acad Sci 2004;1034:284–99. [5] Simpson JL, Elias S, Malinak LR, Buttram Jr VC. Heritable aspects of endometriosis. I. Genetic studies. Am J Obstet Gynecol 1980;137:327–31. [6] Matalliotakis IM, Arici A, Cakmak H, Goumenou AG, Koumantakis G, Mahutte NG. Familial aggregation of endometriosis in the Yale Series. Arch Gynecol Obstet 2008;278:507–11. [7] Autrup H. Genetic polymorphisms in human xenobiotica metabolizing enzymes as susceptibility factors in toxic response. Mutat Res 2000;464:65–76. [8] Hu X, Benson PJ, Srivastava SK, et al. Glutathione S-transferases of female A/J mouse liver and fore stomach and their different induction by anticarcinogenic organosulfides from garlic. Arch Biochem Biophys 1996;336:199–214. [9] Hu X, Srivastava SK, Xia H, Awasthi YC, Singh SV. An alpha class mouse glutathione S-transferase with exceptional catalytic efficiency in the conjugation of glutathione with 7b, 8a-dihydroxy-9a, 10a-oxy-7,8,9,10-tetrahydrobenzo[a] pyrene. J Biol Chem 1996;271:32684–88. [10] Harries LW, Stubbins MJ, Forman D, Howard GC, Wolf CR. Identification of genetic polymorphisms at the glutathione S-transferase Pi locus and association with susceptibility to bladder, testicular and prostate cancer. Carcinogenesis 1997;18:641–4. [11] Wang YF, Zong LL, Mao T, et al. Relationship of polymorphisms of AhR-1661G/ A with GSTP1-313A/G and susceptibility to endometriosis. Zhonghua Fu Chan Ke Za Zhi 2012;47:522–5 [Chinese]. [12] Matsuzaka Y, Kikuti YY, Goya K, et al. Lack of an association human dioxin detoxification gene polymorphisms with endometriosis in Japanese women: results of a pilot study. Environ Health Prev Med 2012;17:512–7. [13] Vichi S, Medda E, Ingelido AM, et al. Glutathione transferase polymorphisms and risk of endometriosis associated with polychlorinated biphenyls exposure in Italian women: a gene–environment interaction. Fertil Steril 2012;97:1143–51. [14] Wu CH, Guo CY, Yang JG, et al. Polymorphisms of dioxin receptor complex components and detoxification-related genes jointly confer susceptibility to advanced-stage endometriosis in the taiwanese han population. Am J Reprod Immunol 2012;67:160–8. [15] Jeon MJ, Choi YM, Hong MA, et al. No association between the GSTP1 exon 5 polymorphism and susceptibility to advanced stage endometriosis in the Korean population. Am J Reprod Immunol 2010;63:222–6.
X. Chen et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 171 (2013) 362–367 [16] Parveen F, Faridi RM, Das V, Tripathi G, Agrawal S. Genetic association of phase I and phase II detoxification genes with recurrent miscarriages among North Indian women. Mol Hum Reprod 2010;16:207–14. [17] Ertunc D, Aban M, Tok EC, Tamer L, Arslan M, Dilek S. Glutathione-S-transferase P1 gene polymorphism and susceptibility to endometriosis. Hum Reprod 2005;20:2157–61. [18] Hur SE, Lee JY, Moon HS, Chung HW. Polymorphisms of the genes encoding the GSTM1 GSTT1 and GSTP1 in Korean women: no association with endometriosis. Mol Hum Reprod 2005;11:15–9. [19] DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88. [20] Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:719–48. [21] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [22] Munafo MR, Clark TG, Flint J. Assessing publication bias in genetic association studies: evidence from a recent meta-analysis. Psychiatry Res 2004;129:39–44. [23] Aban M, Ertunc D, Tok EC, Tamer L, Arslan M, Dilek S. Modulating interaction of glutathione-S-transferase polymorphisms with smoking in endometriosis. J Reprod Med 2007;52:715–21. [24] Huang SX, Wu FX, Luo M, et al. The glutathione S-transferase P1 341C>T polymorphism and cancer risk: a meta-analysis of 28 case–control studies. PLoS ONE 2013;8:e56722. [25] Yu Z, Li Z, Cai B, et al. Association between the GSTP1 Ile105Val polymorphism and prostate cancer risk: a systematic review and meta-analysis. Tumour Biol 2013;34:1855–63. [26] Wang Z, Xue L, Chong T, Li H, Chen H, Wang Z. Quantitative assessment of the association between glutathione S-transferase P1 Ile105Val polymorphism and bladder cancer risk. Tumour Biol 2013;34:1651–7.
367
[27] Ma Y, Wei X, Han G, Xue M, Li G, Li Y. Glutathione S-transferase P1 Ile105Val polymorphism contributes to increased risk of gastric cancer in East Asians. Tumour Biol 2013;34:1737–42. [28] Tan Z, Feng M, Luo Y, et al. GSTP1 Ile105Val polymorphism and colorectal cancer risk: an updated analysis. Gene 2013;527:275–8. [29] Zhao Y, Wang Q, Deng X, Shi P, Wang Z. Quantitative assessment of the association between GSTP1 gene Ile105Val polymorphism and susceptibility to hepatocellular carcinoma. Tumour Biol 2013;34:2121–6. [30] Zhong H, Feng Y, Zheng GX, et al. A meta-analysis of the association between glutathione S-transferaseP1 gene polymorphism and the risk of adenocarcinomas of lung cancer. Cancer Biomark 2013;13:29–35. [31] Liu JJ, Liu JL, Zhang X, Xie L, Zeng J. A meta-analysis of the association of glutathione S-transferase P1 gene polymorphism with the susceptibility of breast cancer. Mol Biol Rep 2013;40:3203–12. [32] Brosens I, Benagiano G, Endometriosis. a modern syndrome. Ind J Med Res 2011;133:581–93. [33] Li Y, Liu F, Tan SQ, Wang Y, Li SW. Estrogen receptor-alpha gene PvuII (T/C) and XbaI (A/G) polymorphisms and endometriosis risk: a meta-analysis. Gene 2012;508:41–8. [34] Caserta D, Maranghi L, Mantovani A, Marci R, Maranghi F, Moscarini M. Impact of endocrine disruptor chemicals in gynaecology. Hum Reprod Update 2008;14:59–72. [35] Caserta D, Bordi G, Ciardo F, et al. The influence of endocrine disruptors in a selected population of infertile women. Gynecol Endocrinol 2013;29:444–7. [36] Caserta D, Mantovani A, Marci R, et al. Environment and women’s reproductive health. Hum Reprod Update 2011;17:418–33. [37] Gentilini D, Perino A, Vigano` P, et al. Gene expression profiling of peripheral blood mononuclear cells in endometriosis identifies genes altered in nongynaecologic chronic inflammatory diseases. Hum Reprod 2011;26:3109–17.